Having said that, I wouldn't attempt to learn a whole topic with an LLM, that would be frustrating.

In fairness, that is a rejection of the actual technology, not the surrounding culture.

But the point about culture of questioning is an interesting one I hadn't heard before.

A good human on the subject will do better, and I'm not claiming the LLM has "understanding" in any philosophical sense. But it's more useful than something that just regurgitates material.

> Po-Shen Loh (Chinese: 罗博深; pinyin: Luó Bóshēn; born June 18, 1982) is an American mathematician specializing in combinatorics. Loh teaches at Carnegie Mellon University, and from 2014 to 2023 served as the national coach of the United States' International Mathematical Olympiad team. He is the founder of educational websites Expii and Live, and lead developer of contact-tracing app NOVID.

(Do I just like arguing? Not really but I really did have no idea what your first comment was about, and your follow up was too odd not to reply to, so I got hooked.)

And you think the front page shouldn't show a basic example because you just vibe code everything anyway. That is a very odd objection - it would be very concerning if a code library didn't include a basic usage example on its front page. Also, guess what, some of us do actually write (or at least look at!) some of the code in the programs we work on. And even AI benefits from a clean API.

And your complaint about a to do list is that it's not "impressive"? It's not meant to be! It's a basic illustration of the API. It's like reading the Rust tutorial and objecting, "ugh, another programming language with for loops".

Is it because it's clichéd? That doesn't matter so long as it gets the point across.

Is it because you object to having desktop applications that do a simple task like this? I remember when simple single-purpose desktop apps were common e.g. Card File in Windows 3.1. Also this is meant to be an example for both desktop and mobile. Are only SPA web pages allowed to do things like this now? What a depressing world if so.

Or is your comment satire? I honestly can't tell.

Edit: What would you allow as a first example of a new UI toolkit. (Not that Flet is new, it's had plenty of 0.x releases.)

I've used this for a fairly complex app at work and it's great. (It's certainly not vibe coded as one commenter suggested.) It's a Python layer on top of Flutter, and it was the only UI framework I could find for Python that was fast and worked on all major desktop and mobile OSs. I think maybe Kivy could too but it seemed like it would be an uphill struggle for a good desktop app.

It's a bit of a gateway drug for pure Flutter. If you need some complex, quickly updating control then you can write a Flutter control (in Dart) and wrap it in a Flet (for Python) control for better performance than doing it all in Python. I've ended up doing that for the main display in our app, and there's always the option to move more and more business logic into that.

One downside to Flet is that I wish they didn't try "fix" some design decisions in Flutter. For example, they have a Flet "Container" widget that wraps a bunch of different Flutter layout widgets. The Flet version does seem easier to understand to me, but there's just so much more documentation (and StackOverflow answers etc.) for Flutter than for Flet that it would be easier overall if they just mirrored the Flutter controls precisely.

   FOR I = 1 TO 10
       FOR J = 1 TO 10
           FOR K = 1 TO 10
               DOSOMETHING I, J, K
               REM Breaks out of the middle loop, continues next I iteration
               IF SOMECONDITION(I, J, K) THEN BREAK J
           NEXT K
       NEXT J
   NEXT I

   def inner_loops(i):
       for j in range(10):
           for k in range(10):
               do_something(i, j, k)
               if some_condition(i, j, k):
                   return
   
   def fn():
       for i in range(10):
           inner_loops(i)

LLMs don't just naturally output a single suggested word (or token) each iteration. Instead, they output a value (roughly, a probability) for every possible word. It seems obvious to simply pick the top (i.e. best) suggestion each time. Then your objection makes sense: watermarking would violate this.

Of course people have tried this! The problem is, in practice this makes the LLM much less "creative" than if you randomly pick one of its suggestions (weighted by the numbers it assigned them). You can artificially increase the value higher-value outputs to reduce the chances of it saying something really odd, and this parameter is called "temperature". A higher temperature allows lower-probability choices (therefore seemingly more creative but perhaps less accurate) and a lower number vice-versa. Either extreme works poorly, and picking a good number is part of optimising an LLM.